Content area
Full Text
1. Introduction
The problem of reflexes of PIE. syllabic resonants in Balto-Slavic is one of the most difficult unresolved problems not only in Balto-Slavic historical phonology, but in Indo-European linguistics as well1. It has long ago been noted that syllabic *r, *n, *m, and *l have two sets of reflexes in Baltic and Slavic, which can be labeled "I-reflexes" (*r > *ir, *l, > *il, *m > *im, *n > *in), and "U-reflexes" (*r > *ur, *l > *ul, *m > *um, *m > *un)2. However, even a hundred years ago it was clear to Endzelins that no simple solution, involving clearly definable environments for different reflexes, can be found: "Ne trudno ved' ubedit'sja, cto bait, ir i t.d., praslav. brit.d. s odnoj storony, i bait, ur i t.d. praslav. br i t.d. s drugoj storony, vstrecajutsja v sosedstve vsevozmoznyx zvukov" (Endzelins 1911: 17). Some linguists, notably A. Vaillant (1950) and J. Kurylowicz (1956: 272-3), have tried to show that there indeed are phonological environments in which U-reflexes are regular, namely, after velars, but their view has not met with general acceptance, like J. Endzelins (1911) and R. Trautmann (1923), have argued that U-reflexes are regular only in roots, whose full grade contains the vowel *o, while I-reflexes are regular where full grade in *e is the norm. This view is easily refuted with such obvious counterexamples as PIE. *g^sup w^hrno- "hearth" > BSl. *gurna- (Russ. gorn, Cz. hrnec "pot"), with e-grade in *g^sup w^hermos "hot" > G. thermós.3
Finally, some linguists, like Y. Shevelov (1964: 87), have resorted to statistics, trying to establish relative frequencies of I- viz. U-reflexes; thus it was claimed that *UR was particularly frequent after labials and velars, while *IR was frequent after dentals and palatals; however, in view of the fact that productive ablaut is certain to have largely erased whatever original distribution there had been, involving statistics should be seen more as a sign of desperation than a way to the solution.
Thus, even the great Baltic scholar Christian Stang had to admit: "hier bleibt ein Problem, das ich nicht zu lösen vermag" (Stang 1966: 80).
It would be wrong, however, to claim that all that has been written so far about the subject was a waste of...